USPSTF adopts decision modeling as one tool to develop recommendations
The U.S. Preventive Services Task Force has adopted decision modeling as one of its methods in developing its evidence-based recommendations about preventive health care.
The USPSTF has developed a framework that will help the independent body of medical experts determine when to use modeling as a complementary, quantitative approach as they deliberate on their recommendations for clinical practice and medical policy.
“This is methodologically important because it describes a framework the Task Force can use to assess when decision models may aid in the development of recommendations,” said Douglas K. Owens, who led the decision-modeling group at the USPSTF.
Owens, of the VA Palo Alto Health Care System, and a professor of medicine at Stanford Medicine and a leader at Stanford Health Policy, said the task force spent three years developing the framework.
“Modeling can be a very useful tool to complement empiric studies and can enable you to answer questions that you couldn’t otherwise answer,” he said. “For example, what’s the implication of starting mammography screening at 40 versus 50, or colorectal cancer screenings at 50 versus 45? How might a preventive intervention perform in a population that is slightly different from the ones that have been studied?”
Modeling provides insights about how benefits and harms of an intervention may vary in such circumstances, he said.
The framework for using modeling was described in the July 5, 2016, online edition of the Annals of Internal Medicine, the journal of the American College of Physicians.
The article emphasizes how decision modeling can be useful in answering important medical questions that have not been addressed in clinical trials.
“Decision models are a formal methodological approach for simulating the effects of different interventions — such as screening and treatment — on health outcomes,” the authors write. “Unlike epidemiologic models that project the course of disease or seek to make inferences about the cause of disease, decision models assess the benefits and harms of intervention strategies by examining the effect of specific interventions.”
The USPSTF has used decision models in the development of recommendations for screening colorectal, breast, cervical, and lung cancer. It’s also used modeling in recommendations for use of aspirin to prevent cardiovascular disease and colorectal cancer.
For these services, decision models allowed the USPSTF to consider the lifetime effect of different screening programs in specific populations and at various ages.
For example, the effect of HIV screening on mortality has not been directly assessed in clinical trials. But there is evidence that HIV can be diagnosed accurately and that early treatment reduces mortality and HIV transmission.
In contrast, a recent assessment of screening for thyroid dysfunction found that the evidence was insufficient to assess benefits and harms, so this topic would not be a candidate for modeling.
Models are also useful for weighing the harms and benefits of an intervention. Clinical studies have shown, for example, that aspirin can prevent non-fatal heart attacks and strokes. But it also causes intracranial and gastrointestinal bleeding.
“How do you weigh these different benefits and harms? Modeling can be very useful in understanding the tradeoff between benefits and harms,” Owens said.
The Task Force is a leader in evidence-based methods for guideline development, Owens noted. “I hope this work will also be useful to the broader community that is developing clinical guidelines.”